Abstract
Despite a proliferation of studies of the micro-level dynamics of protests and petitions against land takings in China, we know very little about how meso-level factors, such as the local economy, influence petitions to Beijing and provincial governments. Drawing upon the economic approach to civil war, this article examines the roles played by grievances and greed in determining the scale of mobilization at the county level in Zhejiang province. Through archival evidence and interviews in Ningbo and Lishui, as well as an original dataset of petitions, this article suggests that both grievances and greed influence petitioning. Mobilization is especially high in Ningbo, where valuable real estate markets have prompted landless farmers to compete with local governments over control of the rents from land. The article proposes the concept of resource value activation as a cognitive mechanism that has contributed to this process of mobilization.
Keywords
Over the past fifteen years land has become the single largest source of conflict in rural China. The micro-level dynamics of protests by landless farmers against local governments have been extensively explored through numerous case studies (Guo, 2001; Hess, 2010, 2015; Heurlin, 2016; Lian, Glendinning, and Yin, 2016; Mertha, 2008). Landless farmers were one of the groups that mobilized extensively during the “high tide” of petitions to Beijing between 2003 and 2006 (Li, Liu, and O’Brien, 2012). Yet scholars have only begun to explore the role of meso-level factors such as local economic conditions in influencing petitions to provincial governments and Beijing during the “high tide” (Chen, 2016a, 2016b) This article attempts to answer the following question: Why did some cities and counties experience high levels of petition mobilization by landless farmers to provincial and national governments while others experienced very little mobilization?
Much of the literature on micro-mobilization in land conflicts suggests that villagers’ grievances—often in the form of economic insecurity—have motivated land protests (Guo, 2001; Walker, 2008; So, 2007; Lian, Glendinning, and Yin, 2016). Paul Collier and Anke Hoeffler (2004) pioneered an approach to explaining the emergence of civil war based on grievances and greed. They operationalized grievances as ethnic or religious hatreds, political repression and exclusion, and economic inequality. By contrast, they operationalized greed as “extortion of natural resources, donations from diasporas, and subventions from hostile governments” (Collier and Hoeffler, 2004: 565). Drawing upon this economic approach to civil war, I suggest that variation in petition mobilization at the meso-level of counties and cities in Zhejiang was also influenced by “greed”: the competition to extract rents from land.
How can grievances and greed be operationalized in the context of land disputes? Grievance-based arguments for protests against land takings typically focus on economic insecurity. Average rural incomes therefore provide a relatively good proxy for the extent of subsistence-based grievances. Yet grievance-based accounts also emphasize that forcible or illegal land takings frequently contribute to land protests (Guo, 2001). The number of violent or illegal land takings or the level of coercion would provide ideal proxies for grievances. Unfortunately, however, such data are unavailable. Instead, I rely on a rougher proxy of grievances: the scale of land transfers. Inasmuch as land takings frequently leave landless farmers aggrieved about compensation or forcible procedures, areas where land takings are more widespread could be argued to have more grievances because there is a larger population of aggrieved landless farmers. Evidence presented below, moreover, suggests that illegal land takings are more common in areas where the scale of land transfers is greater because of scarcity in land-use quotas.
Land values provide a relatively straightforward proxy for greed. In highly valuable real estate markets, villagers often use their land to extract rents through commercial or industrial activities and can thus be thought of more as “landlords” than as “peasants.” In this sense, land values capture greed. I propose that rising land values have resulted in resource value activation, a cognitive mechanism of mobilization in which natural resources that were previously viewed as unimportant become viewed as highly valuable, spurring competition by rival claimants to the resource.
The Role of Grievance and Greed in Explaining Local Variations in Petition Mobilization
At the macro-level, the high tide of petition mobilization to Beijing was driven primarily by an elite split and the turnover in administration from the elitist Jiang Zemin administration to the more populist Hu Jintao administration. The incoming Hu administration’s populist rhetoric spurred expectations that it would be responsive to petitioners’ complaints and prompted a flood of petitioners to Beijing (Li, Liu, and O’Brien, 2012). As Jing Chen (2016a, 2016b) notes, however, geographic variation in the level of mobilization during the high tide and beyond has received much less attention. This topic is important because studies of mobilization over labor and taxation, for example, have found that regional economies exert a strong influence on protests (Lee, 2007; Hurst, 2009; Bernstein and Lü, 2003). Through a county-level dataset from Jiangxi, a relatively poor province, Chen’s pioneering work in Jiangxi shows that more economically developed counties—operationalized as GDP per capita—experienced greater petitioning to Beijing. She hypothesizes that this is due to either the greater resources of citizens in these counties or that their rights are violated with greater frequency. Using provincial-level data, moreover, she confirms that GDP per capita has a positive influence on mobilization but finds that land takings have no impact on petitions to Beijing (Chen, 2016a).
I show that the dynamics of petition mobilization over land in Zhejiang are quite different from Jiangxi. The overall level of economic development in Zhejiang (as measured by GDP per capita) has no impact on petitions to either the provincial government or Beijing, contrary to Chen’s findings in Jiangxi. Farmers’ incomes, moreover, are inversely related to petitions: counties where farmers are poorer tend to experience more petitions. I find that (contrary to Chen’s nationwide provincial-level data on petitions), land takings are a significant driver of petitions to the Zhejiang provincial government and Beijing. Taken together, these results suggest that it is the greater prevalence and severity of grievances, rather than increased resources, that caused local petition mobilization to increase. Equally importantly, I show that local land values influence petition mobilization. My analysis suggests that mobilization is higher not because villagers have more resources, but rather because they are competing with local governments for control over more valuable resources. These differences in patterns of mobilization are likely driven at least in part by differences in the local economies of these two provinces: land is far more valuable in Zhejiang.
What is the mechanism that links local real estate markets to mobilization? As McAdam et al. note, mechanisms “are a delimited class of events that alter relations among specified sets of elements in identical or closely similar ways in a variety of situations” (McAdam, Tarrow, and Tilly, 2003: 24). Collier (2000) has proposed that civil wars frequently occur in resource-rich countries because rebels and the state are competing to extract rents from natural resources. In this sense, rebellion can be considered a “quasi-criminal activity” that is primarily motivated not by grievances with the government, but by greed. Natural resources help to finance the fielding of rebel armies. Scholars have recently suggested that land revenues in China might constitute a “resource curse” (Chen and Kung, 2014).
Drawing upon the economic literature on civil wars and conceptualizing land as a natural resource, I identify a mechanism I call “resource value activation.” Resource value activation operates as a cognitive mechanism. Cognitive mechanisms “operate through alterations of individual and collective perception” (McAdam, Tarrow, and Tilly, 2003: 26). Resource value activation occurs when natural resources that had previously been perceived by actors as relatively unimportant are suddenly perceived as highly valuable. Perceptions of value consequently spur competition over resources between challengers and the state or third parties that wish to exploit the resource. Petitions against land takings—particularly petitions to provincial governments and Beijing—frequently succeed in securing higher compensation (Heurlin, 2016). Yet land values matter precisely because petitioning is risky and costly. The higher the value of the resource over which actors are competing, the greater their willingness to accept the costs and risks associated with mobilization. Just as competition over resource wealth can spark civil wars, it can also provide the motivation for collective protests. Villagers in Peru, for example, have mobilized campaigns demanding greater reinvestment of the proceeds from mining in the local communities from which they are extracted (Arce, 2014).
It is somewhat surprising that petitions should be concentrated in high-value real estate markets. High land values have bolstered government revenues. Recent research has suggested that land protests may occur primarily where local governments lack the funds to compensate farmers (Kuang and Göbel, 2013). Better-funded local governments could be expected to offer higher compensation and potentially have better enforcement of central government policies protecting farmers’ rights in land takings. If subsistence-related grievances are driving disputes, this might similarly suggest that areas with lower real estate values and lower farmer incomes would have more protests. There is some suggestive evidence that this is the case. A comparative study of cities in four provinces found that participation rates in petitions were far higher in the poorer city of Shuangliu, Sichuan, than in middle income Jiaozhou, Shandong, and higher income Beijing and Suzhou, Jiangsu (Han, 2009). Indeed, farmers in high-value real estate markets may even actively collude with local officials to seek expropriation in order to illegally construct rental housing (Paik and Lee, 2012).
There is very little comprehensive data available on the geographic distribution of land protests. Broadly speaking, land values are higher in eastern coastal provinces than in inland provinces. Tong and Lei’s (2014) dataset on land protests based on media reports does not, however, seem to suggest a strong relationship between real estate values and mobilization. Guangdong has by far the highest number of protests (which I suspect may be partially the result of better media coverage), but the second and third highest number of protests were in Yunnan and Sichuan, provinces with far lower land values. Seven of the top twelve provinces for protests were inland provinces. Similarly, in Liu’s dataset of land protests from 2003 to 2011, half of the top ten provinces for land protests were in poorer inland provinces (Liu, 2013). In sum, the limited data available are indeterminate and do not necessarily suggest that mobilization is higher in valuable real estate markets.
Data and Sources
This article draws upon three sources of data. First, it draws on reports from serials produced by the Zhejiang Provincial Land Resources Office entitled Zhejiang guotu ziyuan (浙江国土资源, Zhejiang Land Resources) and by the Zhejiang Provincial Petitioning Office entitled Zhejiang xinfang (浙江信访, Zhejiang Petitions) and Xinfang yu minqing (信访与民情, Petitions and the Condition of the People). The reports are written almost exclusively by land resources and petitioning officials, and frequently analyze the causes of land-related petitions. Second, the article draws upon an original county-level dataset of land-related petitions to the Zhejiang Provincial Land Resources Office and the Ministry of Land Resources in Beijing in 2006. The petition data come from a document I acquired from the Ningbo Land Resources Bureau. Third, the article draws upon interviews I conducted with landless farmers and government officials in counties in Hangzhou, Ningbo, and Lishui conducted in 2009 and 2010. Zhejiang is not an average province. Real estate markets in Zhejiang are much more developed than in most other provinces. Similarly, the evidence suggests that petition mobilization is much higher in Zhejiang than in other provinces. As such, it is clearly not representative of all provinces. Rather, it represents a useful starting point for generating hypotheses on the dynamics of mobilization in valuable markets.
Grievances in Petitions to Land Resources Offices
What do villagers petition about in land-related disputes? If we look at petitions to the Zhejiang Provincial Land Resources Office in 2006, a report summarizing statistics on petitions in the first half of that year notes that there are four types of substantive grievances reported in petitions: illegally occupying land, land-taking disputes, property disputes, and mining disputes (Zhejiang Provincial Land Resources Office, 2006). Illegal land seizures are the most commonly reported grievance, accounting for 40.9 percent of petition letters, 52.2 percent of petition visits, and 48.5 percent of participants in petition visits (see Table 1). This category is relatively broad and includes occupying land without authority, exceeding authority in approving land use, occupying land without obtaining prior approval, seeking approval for less land and taking more, illegally transferring land, and chaotically approving housing land (Ningbo Land Resources Bureau, 2010). The second most common substantive grievance is land-taking disputes, accounting for 19 percent of petition letters, 21.5 percent of petition visits, and 26.5 percent of participants. Evidence from 2004 and 2007 suggests that this distribution of petitions was relatively stable over time (Zhejiang Provincial Land Resources Office, 2005a, 2008). The definition of petitions over land-taking disputes used by the Zhejiang Provincial Land Resources Office is relatively narrow. It comprises two subcategories: (1) refusal to agree with land takings and (2) resettlement compensation (Ningbo Land Resources Bureau, 2011). In practice the difference between these two categories is limited. Importantly, if a violation of land-use procedures occurs in the course of a land taking, it is by definition categorized as an illegal land seizure, not a land-taking dispute. This means that a significant proportion of petitions about illegal land seizures are actually complaints about legal violations in land takings. Indeed, the Ministry of Land Resources frequently does not bother to disaggregate the two categories of illegal land seizures and land takings in its statistics on petitions (Ministry of Land Resources, 2004–2007). Officials in Zhejiang likewise report that the same group of landless farmers often complains both that compensation is too low and that the land taking was illegal (Liu, 2004).
Grievances in Petitions to the Zhejiang Provincial Land Resources Office in the First Half of 2006.
Predatory Land Takings as Grievances
Farmers whose land is taken receive primarily cash compensation. Often, however, the amount of compensation they receive is only 5–10 percent of the land transfer fees paid by the end-user while 60–70 percent goes to the county and township governments, and the remaining 25–30 percent goes to the village collective (Guo, 2001: 428). The majority of landless farmers view this compensation as too low. A survey in Zhejiang in the early 2000s revealed that 53.2 percent of landless farmers felt that the compensation was too low, 22 percent that it was extremely low, and only 23.9 percent felt that it was appropriate (Zhejiang Provincial Rural Social and Economic Research Team, 2003). This low compensation is the basis for grievances centered on subsistence. Grievances might be especially severe in areas where farmer incomes are already relatively low. Petition officials in relatively poor Longyou county (Quzhou), for example, reported landless farmers mobilized petitions associated with underdevelopment, particularly unemployment and a lack of social security programs for middle-aged farmers (Feng, 2002).
Even in the absence of subsistence crises, the predatory and forcible nature of land takings—as well as their sometimes dubious legality—remain significant grievances for farmers. The implications of this for local variations in mobilization are straightforward: areas with more land takings should have more aggrieved landless farmers, some of whom are likely to petition. The archival evidence supports this interpretation. In Jinhua city, for example, officials reported that “the [counties] in which the increase of collective petitions was rather great were all counties where the strength of industrial park construction . . . was rather high” (Fang, Tong, and Chen, 2003: 38).
The petition data from the first half of 2006 also suggest that illegal land seizures were a key grievance. The central government has attempted to control land loss through a system of quotas on land conversion. The system identifies a minimum amount of arable land that lower-level governments must maintain, while at the same time imposing a maximum amount of urban construction land that can be used during a certain period. Land quotas are distributed down along the administrative hierarchy. The central government sets quotas for provinces, which set quotas for cities, which in turn set quotas for counties. As the quota system was intended to curb land loss, quotas were purposefully set lower than the actual demand for land (Lin and Ho, 2005).
Quotas are necessarily quite scarce in cities in Zhejiang with highly active real estate markets that rely on land transfers to generate construction land. According to a “very conservative” estimate by Wang and Tao (undated), although Zhejiang’s quota for the 1997–2010 period was 1 million mu, actual demand was likely as high as 1.4 million mu. Under the land-use plan, the largest quotas for occupying cultivated land for construction purposes went to cities with active real estate markets, particularly the six cities in the Hangzhou-Ningbo-Shaoxing and the Jinhua-Wenzhou-Taizhou regions. Quotas in cities with less active real estate markets were much lower. Yet by 2001—only halfway through the period covered by the plan—the cities with the highest quotas in these two regions had largely depleted their quotas and had begun to “buy” quotas from cities with less developed property markets, particularly Quzhou, Lishui, and Huzhou (Cai, 2012).
Local governments frequently evade quotas and illegally seize land (So, 2007). In Zhejiang in 2006 the legal violation in 90.7 percent of illegal land-use cases was occupying land without prior approval (China Land Resources Yearbook, 2007). According to the Zhejiang Provincial Land Resources Office, “local governments have become the principal agent of illegal land use and illegal land requisitions and takings. . . . The majority of illegal land behavior almost all directly or indirectly have some connection to local governments” (Zhejiang Provincial Land Resources Office, 2005b). Villagers and enterprises are responsible for a minority of illegal land seizures, often with the tacit consent and secret support of local officials. Where quotas are scarce, local governments typically reserve quotas for key projects 重点项目 and large enterprises. This leaves villagers and small enterprises unable to secure quotas and sometimes prompts them to engage in illegal land seizures (Zhang et al., 2009).
While less common, villagers also petition against fellow villagers who have illegally seized land. For example, they petition against villagers who illegally occupy land to build their houses and thereby interfere with the petitioners’ ventilation, sunlight, or water drainage (Zhang, 2015). Petitioning officials report that “already limited residential housing land-use quotas are insufficient to satisfy the demands of households building houses. Consequently, some households building houses switch to illegally occupying farmland to build houses, courtyards and factories . . . causing the masses to petition” (Hu, 2004, 26). Illegal land seizures are a potent grievance because they provide fertile ground for “rightful resistance,” as villagers complain that local governments have violated central policies (O’Brien and Li, 2006). This leads to the following hypotheses:
Hypothesis 1a: Petition mobilization will be higher in counties with lower farmer incomes.
Hypothesis 1b: Petition mobilization will be higher in counties where more land is transferred.
Resource Land Value Activation as Greed
Rising land values have exacerbated conflicts and intensified petition mobilization. Given the difficulty of finding data on land values, this article will use residential housing values as a proxy. The amount of land transfer fees paid by end-users of land is often opaque and may be unknown to villagers until after the land taking, if they are able to acquire the information at all. By contrast, residential housing values are a good proxy for the perceived value of land precisely because they are the most accessible and widely known measure of local land values for residents of an area. In 1994, residential housing prices in Zhejiang were 957 RMB/m2. 1 While both the government and residents began to see new commercial housing as increasingly valuable, farmers perceived farmland itself to be of relatively limited value, principally due to heavy tax burdens. Consequently, mobilization in the mid-1990s was relatively limited. By the time the first round of land contracts expired, some farmers in Zhejiang had even (often informally) given their land to other villagers to farm (Petitions and the Condition of the People, 2007). In the mid-2000s, however, real estate values increased dramatically, and the fast pace of urbanization meant that even farmland—especially on the edges of cities—was becoming very valuable. Average housing prices increased from 957 RMB/m2 in 1994 to 4,510 RMB/m2 in 2006. It was only after housing prices began to increase rapidly that the scale of mobilization jumped, as seen in Figure 1.

Real estate prices and petition mobilization in Zhejiang, 1994–2014.
In 1994, when housing prices were low and farmers did not particularly value land, there was no significant correlation between housing prices and the level of land petition mobilization at the city level in Zhejiang. 2 By contrast, after housing prices rose substantially and farmers began to compete over land, there was a significant correlation between housing prices and petitioners at the city level in 2005. 3 These results can be seen in Table 2.
Housing Prices and Petitioners in Zhejiang.
Note. Pearson correlation coefficients shown. N = 11.
Source. Author’s calculation based on data from the Zhejiang Land Gazette and the Zhejiang Statistical Yearbook 1995.
p < .05.
Rising land values have transformed the countryside and exacerbated conflicts in four ways. First, rising land values have greatly outpaced land-taking compensation in highly valuable real estate markets, making even land takings of rural farmland unpalatable to many landless farmers. In late 2002, the Zhejiang provincial government introduced the “unified area price” compensation method, which established standard compensation rates for specified areas within each county. The compensation considered not only the output value of the land but also the location, type of land, average landholdings of farmers, and the level of economic development. This new compensation method increased minimum compensation levels by at least 50 percent and in some areas compensation nearly doubled (Heurlin, 2016). While this reform significantly increased compensation, the method still relied on administratively set compensation amounts, meaning that the legacies of the planned economy compensation system had not been completely eliminated. Moreover, compared to the wide variation in land values in Zhejiang, there is relatively little variation in minimum compensation standards. In 2014, for example, the lowest minimum compensation was 37,000 RMB per mu, while the highest minimum compensation was only 54,000 RMB (Zhejiang Provincial Government, 2014).
Second, rising land values caused a boom in rural house construction and generated opportunities for rental income. In the early 2000s, Zhejiang villagers viewed new housing—unlike agriculture—as a good investment and a way to further their demographic and social aims (Sargeson, 2002). Villagers in more developed real estate markets—especially in “villages inside the city” 城中村 (chengzhongcun) and villages on the urban-rural fringe 城郊村 (chengjiaocun)—have often capitalized on land values by constructing housing to rent out, especially to migrant workers. These villagers have become much more reliant on rental income than farming, “turning agriculturalists into experienced and competitive landlords” (Zhao and Webster, 2011: 531). The rising value of land, however, meant that local governments began seizing not only farmland, but also farmers’ housing land. In comparison to urban housing demolitions on state-owned land, moreover, land takings that cause the demolition of housing on rural collective land are far less regulated and offer far less compensation (Shao 2009; Ningbo Petitioning Bureau Petition Reception Office, 2005). Farmers who have rental incomes that will be lost if a demolition occurs are much more willing to bear the costs of mobilizing petitions in order to bargain for higher compensation. These dynamics have caused petition mobilization to be geographically concentrated in chengzhongcun and on the urban-rural fringe (Fang, 2008).
Third, as land values increased and rural industrialization proceeded, villagers realized they could make much more off of their land by (often illegally) building workshops and factories. Lacking access to land, private entrepreneurs frequently used the expansion of their houses as an excuse to build additional workshop space (Whiting, 1999). Unfortunately, however, compensation policies for commercial and industrial-use land are underdeveloped and do not offer substantial compensation for their forgone income. As a result, villagers who previously had income from commercial and industrial ventures tend to lose more potential future earnings than villagers who only had agricultural or rental income (Zhao and Webster, 2011).
Fourth, and relatedly, rising land values also allowed village collective organizations to develop the collective economy. Compensation for collectively owned township and village enterprises is also quite low. The loss of stipends from collectively owned township and village enterprises following land takings frequently motivates landless farmers to participate in petitions (Zhang, 2014). In highly valuable real estate markets like Shangcheng, Hangzhou, demolition of collective industrial and commercial assets became one of the main causes of petitions (Lu, 2009). Making matters worse, local governments often do not report to higher levels in their land-use plans that the land upon which rural factories are built is no longer cultivated land. As a result, when land takings occur, the compensation for the land is based on formulas for farmland instead of construction land, resulting in even lower compensation (Zhuo, 2002).
As the value of retaining land became activated in the minds of many farmers, they became much more willing to engage in costly and risky collective action in order to bargain for higher compensation. As an official from Shangcheng, Hangzhou—one of the most developed real estate markets in Zhejiang—explained, The government controls land-taking prices at a low level, and the market housing price is high. The profits are rich and generous. Peasants can’t get over this. . . . When land is in the hands of farmers, they only get a couple of tens of thousands or maybe 100,000 RMB per mu, and, moreover, after the land goes through a government land taking and demolition and is invested in house property rights, the land value rises suddenly and sharply to a million or even several millions. Farmers’ psychology has no way to reach an equilibrium. (Lu, 2009: 25).
Rising land values caused farmers to eventually see land takings and seizures as an opportunity for enrichment through petition mobilization. Grievances were complemented and sometimes replaced by greed. As a petitioning official explained, “along with the rapid socioeconomic development, the value of land has been constantly increasing. Farmers’ fever for land has been reignited,” causing an increase in petitions. Farmers shifted from an “uncaring” 无所谓 attitude about land rights to a “fight for every inch of land” 寸土必争 attitude (Petitions and the Condition of the People, 2007). Moreover, by the early 2000s, farmers began to see local governments as responsive to protests against land takings, often offering more compensation to farmers who petitioned to higher levels of government (Heurlin, 2016). One official reported that “at least in the east and in the suburbs of large cities, land-taking and demolition compensation have already become the way that locals hope to get rich quick by foul means” (Mei, 2013: 1). Most tellingly, internal reference reports in Zhejiang noted that “the majority of the masses who are not dissatisfied with land-taking compensation policies or are only somewhat dissatisfied also participate in petitions” (Zhang and Zhang, 2004: 37 [emphasis added]). Petition mobilization was driven less by grievances than greed. This leads to the following hypothesis:
Hypothesis 2: Counties with higher land values experience more petition mobilization.
Greed and Grievance in Ningbo and Lishui
These dynamics can be illustrated through brief case studies of two cities in Zhejiang: Ningbo and Lishui. According to the data I collected from the Ningbo Land Resources Department, the average city in Zhejiang sent 70 petition visits to the Provincial Land Resources Office in 2006, with an average of 173 total participants. Mobilization in Ningbo was considerably higher, with 118 visits and 325 participants. Mobilization in Lishui, by contrast, was relatively limited, with only 42 petition visits and 156 participants. Even accounting for the size of their agricultural populations, mobilization in Ningbo was more extensive, at 3.2 petition visits and 8.7 participants per 100,000 residents, versus 2 petition visits and 7.4 participants per 100,000 residents in Lishui. Fully half of counties in Ningbo were in the top third of mobilization in terms of participants per capita, compared to only 22 percent of counties in Lishui, while no Ningbo counties were in the bottom third of lowest mobilization, compared to 33 percent of counties in Lishui.
Ningbo
Land markets in Ningbo were the most active of all of the cities in Zhejiang. At least 926 hectares of land were transferred in Ningbo in 2006 (data from two of Ningbo’s counties was missing). The scale of land transfers was twenty-two times higher than in Lishui. The average county transferred almost 97 hectares of land, more than twice as much as in all of Lishui. 4 One land resources official estimated that land takings generated two thousand landless farmers in Ningbo every year (Xie, 2014). Grievances in Ningbo, in other words, were far more widespread than in Lishui.
The large scale of land takings caused Ningbo to quickly exhaust its supply of land-use quotas. Illegal land seizures, the mayor admitted, were an “extremely serious” problem in Ningbo (Liu, 2011). In 2006, the Ningbo Land Resources Bureau investigated 2,651 cases of illegal land use (Ningbo Land Resources Bureau, 2006b). The scarcity of land quotas, combined with the lengthy approval process, caused county governments in Ningbo to violate the Land Management Law by conducting the land taking at the same time that they sought approval or not even seeking approval at all. These illegal land takings caused a constant increase in mass petitions (Jin, 2008).
The lack of quotas meant that many—perhaps most—houses were illegally built. Higher property values pushed migrant workers out of the urban housing market, at the same time increasing demand for rental housing on rural collective land. This in turn prompted Ningbo farmers to illegally construct rental housing (Fenghua City Government Office Research Group, 2007). Illegal housing increased petition mobilization in Ningbo in two ways. First, illegally built houses caused petitions. One landless farmer I spoke with in Ningbo admitted to having illegally constructed her house, violating restrictions on the height of the building. Her neighbors petitioned and she was forced to pay a fine (Interview 40). Second, illegally constructed housing also made petitions against land takings more likely. In Ningbo, over 70 percent of the houses demolished had been illegally constructed (Ningbo Land Resources Bureau, 2006b).
High real estate prices only exacerbated disputes in Ningbo. Within Ningbo, petition mobilization was generally highest in the areas with higher land values, especially “urban villages” and on the urban fringe 周边, and lower in the more remote rural areas where land values were lower (Interview 182). By 2006, average housing prices in Ningbo reached 5,438 RMB/m2, higher than in all other cities except Hangzhou and Wenzhou. Under the planned economy-era system for administratively setting land-taking compensation, on average landless farmers receive 5–10 percent of the proceeds of a land taking, an unfair distribution that is at the core of many disputes. High land prices have the effect of increasing the disparity between compensation and the auction prices for land. Even Ningbo officials admitted that while the compensation standards were within legal limits, the disparity with land transfer prices was huge (Chen, 2010). The highest tier of compensation was 96,500 RMB per mu, but that same mu of land could fetch over 10 million RMB, meaning that the landless farmers received less than one percent of the proceeds. Officials clearly understand that these land values shortchange farmers. As a land resources official asked wryly, “No matter how you explain this problem, you tell me, can a farmer feel balanced in his heart?” (Xie, 2014: 31). Instead, “some landless farmers become dissatisfied because they cannot objectively understand the reasons for the increase in the auction price; they think the resettlement compensation is too low, and this causes conflicts” (Xie, 2009, 46).
Even in cases in which landless farmers might not fully realize the value of their land, this information is revealed to them at the time that the land-use rights are sold. This can cause spontaneous resource value activation and encourage mobilization. In one memorable example, farmers in a Ningbo village received 20,000 RMB per mu as compensation and were apparently satisfied because they did not petition. After the land taking was completed, the government auctioned off the land for 800,000 RMB per mu, meaning the landless farmers received only 2.5 percent of the proceeds. The very next day over a hundred farmers from the village collectively petitioned the municipal government to demand higher compensation (Xie, 2014).
The rapid growth in real estate prices further exacerbated problems, as administratively set compensation rates struggled to keep pace with market forces. A landless farmer I spoke with in Ningbo attributed disputes in his village to this very dynamic. He noted that compensation rates were always based on land prices in the preceding year, which caused problems in the context of rapidly rising prices (Interview 38). Rapidly rising land prices also reopened old conflicts. Landless farmers who had been relatively satisfied with their compensation when their land was taken three to five years earlier were beginning to mobilize protests because land values had risen so dramatically in the interim (Interview 39).
The high land values also caused the differences in compensation levels between projects to diverge sharply. Comparable land might be compensated at only 6,000–8,000 RMB per mu for highway construction projects, but would commonly be tens of thousands or even a hundred thousand RMB per mu for non-infrastructural land takings. Infrastructural projects therefore became a major source of contention as farmers protested the low compensation relative to other nearby projects (Jin, 2008).
Rising land values, moreover, transformed the structure of land use and the village economy in a manner that generated further mobilization. Farmers shifted their land away from agricultural uses. This meant that when land takings occurred, the claims were far more complex than the low-compensation grievances commonly reported in the early literature on petitions against land takings. Indeed, many Ningbo “farmers” found that farming was not profitable enough to be worth their time. Unable to sell the land, they rented it to migrant workers. When this farmland was taken, the native villagers—not the migrants who farmed the land—received the lion’s share of the compensation. This caused the migrants to petition over the unfair distribution of compensation (Interview 171).
The presence of numerous migrant workers, moreover, opened opportunities to derive significant incomes from rental housing, especially in chengzhongcun in the Jiangdong district (Chu and Chen, 2013). Compensation rates for housing on rural collective land that was demolished, however, lagged behind market prices by 30–40 percent. Moreover, this compensation formula did not take account of the fact that—unlike urban residents whose houses were on state-owned land—villagers held housing construction property rights to the land their houses stood on. Even these compensation rates, moreover, only applied to houses within the urban planning area, and not to large infrastructure projects (Ningbo Land Resources Bureau, 2006a).
When demolitions were imminent, Ningbo villagers attempted to game the compensation system by hastily assembling illegal structures to maximize their compensation by increasing their square footage (Interview 44). This left local officials in a bind: if they refused to compensate landless farmers for these structures, it spurred further resistance, but if they awarded compensation, it only encouraged further illegal construction (Ningbo Land Resources Bureau, 2006a). In one district, officials tried to sidestep this problem by compensating structures built at least ten years earlier at a higher rate than those built more recently, but lamented that it was difficult to ascertain building dates. “Who can keep all these things straight?” the official complained (Interview 171). Villagers even faked divorces and marriages to increase their eligibility for compensation. One Ningbo villager had her hukou in another village and her husband had a non-agricultural hukou. To become eligible for compensation, she divorced her husband and married his father, who did have a hukou in the village (Radio Free Asia, Mar. 22, 2013). These commercial interests and the large amounts of money at stake made the Jiangdong chengzhongcun an epicenter of fierce resistance to land takings (Chu and Chen, 2013).
Many Ningbo villagers had also moved into the service sector, opening small shops in their houses. At all of the villages I visited in one city under Ningbo, villagers reported that most of the villagers who petitioned during recent land takings had been shop owners, while most of the farmers had not petitioned (Interview 38). These shop owners sometimes demand in their petitions that they be given a shopfront in an equally valuable location (Interview 40). Many of the shops were illegally built, further sparking disputes. One shop owner I spoke with was so enraged by the local government’s refusal to compensate her for her illegally built shop that she went to demand an audience with the head of the Urban Construction Department in Ningbo. After the guards refused to let her into the municipal government compound, she revved up her motorcycle and crashed through the gates, whereupon she was promptly arrested (Interview 43).
The rural industrialization that drove increases in land values also allowed villages in Ningbo to develop their collective economy. Most of the collective’s income came from renting factory space to firms. Once demolished in land takings, however, the factory rental space was rarely rebuilt, resulting in the loss of considerable collective assets (Jiangbei Government, 2002). The scarcity of land quotas in Ningbo meant that even if a quota was allocated, it would typically be far outside of the city center, in some cases as far as 40 kilometers (Huang et al., 2009). Property rights to collective assets are also generally unclear, causing landless farmers to fear that they would miss out on compensation (Jiangbei Government, 2002). These dynamics meant that the demolition of factories was particularly prone to cause petitions.
In sum, the consequence of the large amount of land transferred and the valuable real estate market was a very high level of mobilization in Ningbo. Incomplete statistics from Ningbo’s three core urban districts (Haishu, Jiangdong, and Jiangbei) show that there were 363 petitions involving 2,395 landless farmers whose houses were demolished between 2000 and 2006. Of these, there were 14 collective petitions that involved over fifty landless farmers, with an average of seventy-three participants (Ningbo Petitioning Bureau, 2006).
Lishui
Compared to Ningbo, land values in Lishui were not nearly as high. The average housing price in 2006 was 4,405 RMB/m2, just below the provincial average. Property values had, however, increased considerably in the past few years, rising from only 2,475 RMB/m2 in 2002. The scale of land transfers was also limited. In 2006 only 42.7 hectares of land were transferred in Lishui. The average county transferred just over 3 hectares. 5 Real estate values were especially low in my field site in Lishui—a rather remote county I call Songyuan. Average housing prices there were only 3,589 RMB/m2.
The relatively inactive real estate markets meant that, unlike in Ningbo, land quotas were not particularly scarce. Indeed, Songyuan had a surplus of land-use quotas each year, which it sold to raise funds (Interview 183). Consequently, illegal land seizures were a far less common grievance in Lishui. Lishui city investigated only 907 illegal land-use cases in 2007, compared to 2,651 in Ningbo in 2006 (Ningbo Land Resources Bureau, 2006b; Lishui Land Resources Bureau, 2007).
By and large, however, only two or three households in each village built houses illegally (Interview 194). Illegal housebuilding, in particular, was largely confined to the county seat, where land values were somewhat higher (Interview 159). As in Ningbo, Songyuan officials were hesitant to penalize villagers by demolishing their illegal buildings—which generally caused them to petition—and instead tended to levy fines (Interview 193). A land resources official estimated that roughly 10 percent of illegal housebuilding cases escalated into petitions when they attempted to punish the violators (Interview 194). Illegal land takings by local governments in Lishui, however, did sometimes cause petitions (Hu, 2004).
Weak real estate markets meant that there was relatively little competition over land. In Songyuan, for example, officials reported that where land values were especially low—particularly in outlying or mountainous parts of the county—landless farmers were not very opposed to land takings (Interviews 194, 159). Petitions were instead concentrated primarily in the county seat, where land values were higher (Interview 159). The deputy county party secretary attributed the higher level of mobilization in the county seat directly to greed: “It’s a problem of money. . . . It’s an economic problem” (Interview 115). Overall, however, local officials regarded the level of mobilization compared to the rest of Zhejiang as relatively limited (Interview 115).
The structure of the rural economy was much more agricultural, resulting in a far less complex and varied landscape of petition mobilization. Only 30 percent of villagers in Songyuan had off-farm jobs (Interview 184). This meant that Songyuan not only avoided many of the petitions related to industrial and commercial interests but also caused some subsistence-related grievances. A Songyuan land resources official reported that many able-bodied farmers had already migrated out of the county, leaving behind only the elderly, who had difficulties finding jobs after land takings, worried about their future subsistence, and therefore petitioned (document on file with the author). According to a Songyuan petitioning official, land-taking disputes centered largely on three problems. First was compensation for land. Until 2003, land-taking compensation was relatively low and varied within a narrow range, with officials generally giving compensation on the lower end of the range. In 2004, the unified area price was implemented and compensation levels increased. Local officials also began giving compensation on the higher end of the range. The official claimed that landless farmers were much more satisfied than they had been previously. Second was the distribution of compensation within the village. In particular, most petitions concerned women who married in or out of the village and university students who had transferred their hukou to the city where their university was located. Originally Songyuan officials gave university students half the compensation, but in 2007 began offering full compensation (Interview 159). Soldiers who are promoted often transferred their hukous, only to regret this decision when a land taking occurred and they were denied compensation (Interview 115). Finally, many disputes were indirectly related to land takings, such as irrigation drainage problems after a land taking or a road that was paved but not all the way to the resettlement house (Interview 159).
The pattern of conflict that emerged was a low level of mobilization driven primarily by sporadic grievances in the presence of the occasional land taking. Low property values meant that resource value activation occurred to a much more limited extent, primarily in the county seats, and that greed was much less of a factor in mobilization. A Songyuan petition official reported that the county experienced only seven to nine land-taking disputes between 2008 and 2009 (Interview 159). A land resources official, however, reported that there were roughly a hundred petitions to the land resources bureau each year, most of which complained about land takings, illegal land seizures, and illegal housebuilding (Interview 194).
County-Level Variation in Zhejiang
The availability of data for all of Zhejiang’s 90 county-level jurisdictions in 2006 provides the opportunity to conduct a more rigorous statistical examination of the impact of real estate markets on the mobilization of land-related petitions. I use two dependent variables: first, the number of petitioners in petition visits to the Zhejiang Provincial Land Resources Office in Hangzhou; and second, the number of petitioners in petition visits to the Zhejiang Provincial Petitioning Office in Hangzhou and the Ministry of Land Resources in Beijing. Unfortunately, the data did not disaggregate between the Petitioning Office and the Ministry of Land Resources. As such, both dependent variables are best understood as measuring the extent of petitioning to “high level” governments, including provincial governments in the case of the former and the provincial and central government in the case of the latter.
The highest mobilization in terms of both petitions to the Petitioning Office and Ministry of Land Resources and the Zhejiang Land Resources Office was in Yuyao city, Ningbo, at 150 and 181 petitioners, respectively. By contrast, no petitioners to Beijing or Hangzhou came from Chun’an (Hangzhou) and Dongtou (Wenzhou) counties. This illustrates that even high mobilization cities such as Hangzhou and Wenzhou were home to counties with extremely low mobilization. On average, the number of petition visits to the Land Resources Office was roughly ten, with an average of roughly 33 petitioners from each county. What is most notable about these figures is that petition visits are surprisingly small in terms of participants, with just over three participants in the average collective petition.
In terms of the petitioners to the Zhejiang Land Resources Office in 2006, petitioners were highly concentrated in several county-level jurisdictions. The ten jurisdictions of Yueqing city, Yiwu city, Lucheng district, Yongjia county, Cixi city, Huangyan district, Wenling city, Dongyang city, Wencheng county, and Rui’an city accounted for 30.8 percent of all petition letters. Notably, with two exceptions, all of these jurisdictions were more urbanized county-level cities or districts, and not counties, which tend to be more agricultural. In regional terms, all of the jurisdictions were in relatively developed regions of Hangzhou-Ningbo-Shaoxing and Wenzhou-Taizhou-Jinhua, and five of the ten jurisdictions were in Wenzhou. In terms of petitioners to the Petitioning Office and Ministry of Land Resources, districts and county-level cities likewise accounted for 80 percent of the top ten jurisdictions for petitions to the Petitioning Office and Ministry of Land Resources. In regional terms, mobilization was highest in the Hangzhou-Ningbo-Shaoxing region. None of the top ten county-level jurisdictions were in Wenzhou and only one (Jiaojiang district) was in Taizhou. Notably, two of the top ten jurisdictions (Longquan and Qingtian) were in Lishui. The geographical division of petitioners to the Land Resources Office is displayed in Figure 2. What explains this regional variation? In particular, how do local real estate markets influence land-related petitions?

Petitioners to the Zhejiang Provincial Land Resources Office in 2006.
Most of the focus of the previous sections has been on land takings and illegal land seizures by local governments and (to a lesser extent) villagers. Together, these two categories account for 75 percent of all the petitioners to the Land Resources Office. At the county level, however, the petitions are not disaggregated by cause. This represents a limitation of this study, as it is not clear if the distribution of grievances in petitions is relatively consistent across counties. Additionally, it is not clear how real estate markets might influence advice or criticism petitions. Despite these limitations, however, the dataset represents the best available data for gauging the impact of the local economy on land-related petitions.
I operationalize grievances in two ways. First, and most directly, I examine the net per capita income of rural residents. This can be conceptualized as a proxy for the “severity” of subsistence-based grievances. Second, I use the scale of land transfers. This measure of grievance is less straightforward. The discussion above suggested that illegal land takings were more common in highly active real estate markets due to the scarcity of land conversion quotas. The scale of land transfers can be conceptualized as a (very rough) proxy for the “prevalence” of grievances. As the scale of land takings increases, there are more and more potentially aggrieved landless farmers who can mobilize petitions. The Ministry of Land Resources’ Land Transaction Monitoring System (www.landchina.com) provides comprehensive data on all land transfers at the county level. I coded every single transaction in each of Zhejiang’s districts and counties in 2006. Notably, the land-transfer data include both takings of rural collective land and transfers of urban state-owned land. Data from fifteen county-level jurisdictions were missing. Despite this limitation, the data are the best available on land takings.
I operationalize greed by focusing on commercial housing prices. Housing prices provide a straightforward measure of land values and a commonly used benchmark of value by landless farmers. Finally, I attempt to control for several factors. GDP per capita captures the overall level of development in a county. In order to help capture variations in the quality of local governance, I control for local court funding. 6 Higher court funding may be partially a result of local state fiscal resources. Controlling for GDP per capita helps mitigate this problem, but it remains a very rough proxy for the quality of local governance. Finally, the regression controls for the size of the agricultural population, as more populous counties are likely to mobilize more petitions. 7
Results and Discussion
In order to examine the determinants of petition mobilization at the county level, a negative binomial regression is used. This model is appropriate because the data are highly skewed toward few or no petitioners. Model 1 provides initial support for the greed hypothesis: counties with higher average commercial housing prices tend to have more petitioners to the Provincial Land Resources Office. Model 2 also provides initial support for the grievance hypothesis: counties where more land is transferred tend to have more petitioners to the Land Resources Office (see Table 3).
Negative Binomial Regression on Petitioners to Zhejiang Provincial Land Resources Office.
Note. Coefficients given with robust standard errors in parentheses. Model 1: N = 89; Model 2: N = 75; Model 3: N = 67; Model 4: N = 51; Model 5: N = 67; Model 6: N = 52.
p < .01. **p < .05. *p < .1
Models 3 through 6 add additional controls and use alternate specifications of the key independent variables. Aside from the occasional missing case, comprehensive data for counties, county-level cities, and districts are available for all of the variables, with the exception of housing prices and net rural income per capita. For housing prices, comprehensive data are only available for counties and county-level cities (but not districts). However, aggregated data are available for “districts under a prefectural city.” For example, for Qujiang district and Kecheng district, the dataset uses the housing prices for districts under Quzhou city—the prefectural-level city that administers the two districts. Since values are given on a per square meter basis, the values are comparable to the county and county-level city values despite their higher level of aggregation. By doing this, the dataset avoids losing data for most of the 33 districts. Models 1 and 3 use the measures for the housing prices that include data from these districts. In order to show that the relationships hold even if these missing data from districts are excluded, Models 4 and 6 replicate the results without the districts.
As Model 3 shows, average housing prices positively influence the number of petitioners even when controlling for several other factors. Land transfers—which might be conceived of as representing the prevalence of grievances—also have a positive impact on mobilization. The overall level of economic development—as measured by GDP per capita—has no impact on petitions, but court spending has a negative impact on petitions. Although court spending may be more of a sign of wealth than commitment to the rule of law, including GDP per capita as a control variable helps to mitigate this problem. Given the blunt nature of the proxy, however, it is not wise to draw too many conclusions. Model 4 shows that the key explanatory variables remain significant even when districts without housing price data are excluded. Two of the controls, agricultural population and court spending, are no longer significant. Court spending may no longer influence petition mobilization because court spending tends to be highest in urban districts, some of which also have relatively low mobilization.
Models 5 and 6 include rural income per capita, a measure of the severity of subsistence-related grievances. Unfortunately, data for net rural income per capita were not available for each of the districts in Ningbo and Hangzhou. However, much as with the housing price data, aggregated data were available for “districts under Ningbo” and “districts under Hangzhou.” Model 5 uses the aggregated data for these districts. Rural net incomes are insignificant in this model. Model 6 excludes all of the districts with missing data on either rural net income or housing prices. The results of Models 5 and 6 provide support for the greed hypothesis that land values influence mobilization. The results provide mixed support for the grievance hypothesis. The scale of land takings—a proxy for the prevalence of grievances—continues to influence petitions in both models. Rural household incomes—a proxy for the severity of subsistence-based grievances—influence petitions only in Model 6. It is possible that this is because of the aggregated data on incomes in Hangzhou and Ningbo in Model 5. The negative impact of rural household incomes in Model 6 is consistent with the archival evidence, suggesting that livelihood problems drive petitions by landless farmers in some poorer cities in Zhejiang. Rural incomes are also positively correlated with real estate values and land transfers. Since mobilization is highest in cities such as Ningbo, Hangzhou, and Wenzhou—which boast the strongest and most active real estate markets—this suggests that the effects of real estate markets are more powerful than rural incomes in Zhejiang.
Table 4 replicates the results using the number of petitioners to the Zhejiang Provincial Petitioning Office and the Ministry of Land Resources in Beijing as the dependent variable. Notably, the measures for grievances and greed—land transfers and housing prices, respectively—also influence petitions to Beijing. Much like previously, rural household incomes—an alternate measure of grievances—are only significant in Model 6.
Negative Binomial Regression on Petitions to Ministry of Land Resources in Beijing.
Note. Coefficients given with robust standard errors in parentheses. Model 1: N = 89; Model 2: N = 75; Model 3: N = 67; Model 4: N = 51; Model 5: N = 67; Model 6: N = 52.
p < .01. **p < .05. *p < .1.
The quantitative data are consistent with the hypotheses that petitions are driven by both grievances and greed. However, it is important to qualify that the proxies for greed and grievance remain rough and are difficult to untangle given the available data. Until finer-tuned data on illegal land takings and land-taking compensation become available, the quantitative evidence remains exploratory. The quantitative data are, however, consistent with the qualitative evidence presented above that areas with more land takings and higher land values experience higher mobilization.
Does this pattern of grievances and greed influencing mobilization hold beyond Zhejiang? Ultimately this question must be left for future research. As an initial plausibility probe, however, I collected data on the number of participants in land petitions in 2010 from municipal-level yearbooks and land resources yearbooks. More information on this dataset is available in the Appendix. The number of petitioners was unreported in the majority of yearbooks, but fifty-seven yearbooks did provide such information. This represents slightly less than a fifth of all Chinese municipalities. The results of a negative binomial regression (see Table 5) offer more support to the greed hypothesis than the grievance hypothesis. The measures for the severity of grievances (net income per capita) and the prevalence of grievances (scale of land transfers) do not influence the number of petitioners. By contrast, the proxy for greed (land revenues per hectare) does influence the number of petitioners. Cities with higher land values experienced higher levels of mobilization. This holds true for Model 1, which includes cities in Zhejiang, and Model 2, which excludes Zhejiang cities. Although far from conclusive, the evidence is consistent with the hypothesis that high land values are spurring petitioning.
Negative Binomial Regression on Land Petitioners in Cities in China.
Note. Coefficients shown with robust standard errors in parentheses. Model 1: N = 57; Model 2: N = 54.
p < .01. **p < .05. *p < .1.
Coercion, Petitioning, and Real Estate Markets
Given the emergence of a “security state” (Wang and Minzner, 2015) at the local level in China, what role does coercion play in this process? Unfortunately, comprehensive data on public security expenditures at the county-level are unavailable. Coercion against petitioners to high levels of government most frequently takes the form of petition interception, when local government officials detain petitioners en route to the provincial capital or Beijing. Yet petition interception is extremely expensive, with retrieving petitioners to Beijing costing as much as 10,000 RMB each time (Interview 264). Especially during sensitive periods—such as sessions of the National People’s Congress—local officials are frequently stationed at train stations to intercept petitioners (Interview 11). Given the high costs in manpower and resources, it seems likely that better-funded local governments would be able to mount more effective petition interception operations. Local governments, moreover, receive a significant amount of their revenues from land transfers. Assuming all of this is true, the level of mobilization in valuable real estate markets may be even higher than the number of petitions registered at the Ministry of Land Resources in Beijing would suggest. This is necessarily speculative but suggests that high land values may also aid local governments in suppressing petitions.
Conclusion
This article has shed light on a subject that remains understudied in contentious politics in China and in land-related contentious politics in particular: How do local economic conditions influence the mobilization of petitions to higher-level governments? Research from Jiangxi on petitions to Beijing shows that mobilization is higher in more developed counties. An open question, however, has been whether mobilization is higher because more developed counties suffered more frequent violations of residents’ rights and interests, or if mobilization is higher because residents had greater resources (Chen, 2016a). In this article I have offered first steps toward disentangling these two factors through an analysis of Zhejiang. Although the proxy for rights violations (the scale of land transfers) is rough, the analysis suggests that rights violations result in higher mobilization. By contrast, the level of resources (farmers’ incomes) is negatively correlated with mobilization. This finding is consistent with the subsistence-based grievances that earlier case studies have uncovered. Because the proxies remain relatively rough, however, this result is exploratory rather than definitive.
More importantly, however, this article has demonstrated through archival, interview, and quantitative evidence that mobilization of petitions to higher levels of government is also higher in more valuable local real estate markets. Drawing upon the economic approach to civil war, I suggest that local variations in petition mobilization have in this sense been driven by greed. Proposing a mechanism of resource value activation, I have argued that rising real estate values have increased mobilization because villagers can extract greater rents from their control over the land, making them more willing to bear the costs and risks of mobilization.
Several caveats are in order. First, these findings represent a snapshot of mobilization in 2006 at the tail end of the “high tide” of petitioning to Beijing. This was a period of heightened repression because local officials were under severe pressure from the central government to limit petitions to Beijing (Li, Liu, and O’Brien, 2012). Should more data become available, future research should examine the extent to which real estate markets influence mobilization in subsequent time periods when the pressure to prevent petitions was lower. At the provincial level, however, petition data from Zhejiang suggest that mobilization continued to move in tandem with land values in the early 2010s. Housing prices and land petitions in Zhejiang rose through 2012. In 2013–2015, housing prices remained basically stable while petitions began to decline (Heurlin, 2019). This might be consistent with a weakening of resource value activation as real estate markets cool off.
Second, the data are confined to a single province. Zhejiang is a highly developed province where the scale of land transfers has been very widespread and real estate markets—especially around Hangzhou and Ningbo—are very hot. While this provides an important counterweight to the existing study of Jiangxi (Chen, 2016a), future research should examine the relationship between local economic conditions and mobilization of petitions to high level governments in other provinces, particularly in central China. The dataset of municipal-level petitions in 2010 suggests that high land values spur mobilization in other provinces, but future research should test this hypothesis more rigorously through time series data.
Footnotes
Appendix
Housing price data for districts were only available for all districts under a particular prefectural-level city. Simply dropping these cases would result in significant data loss, as there are more than thirty districts in Zhejiang. By using primarily independent variables that can be measured in percentages or on a per capita basis, I was able to retain these cases. The result, however, is that for some variables, districts under the same prefectural-level city all share the same values. Seven of the eleven prefectural-level cities have two or fewer districts, while Taizhou and Wenzhou have three each. The variation between these districts is likely small. Ningbo and Hangzhou, however, have six and eight districts, respectively, meaning that there is likely more variation among these districts. Even when dropping the districts, however, housing prices maintain their statistical significance in all models. Most of the missing data on land transfers are for core urban districts.
I collected the 2010 municipal-level petition data based on municipal and land resources yearbooks in the holdings of the Chinese University of Hong Kong, the Chinese National Library, and the Apabi yearbook database (http://apabi.com/nlc/?pid=yearbook.index&cult=CN). I chose the year 2010 because the data were collected as part of a separate project on the 2010 census. Data on land transfers and land revenues per hectare come from the China Land Yearbook, which reports data for municipalities (but not county-level jurisdictions). Data on my previous measure of land values—residential housing prices—were not available in comprehensive form at China Data Online, so I used the data on land revenues per hectare.
Acknowledgements
The author thanks the editors, referees, Jing Chen, Xi Chen, Martin Dimitrov, Manfred Elfstrom, Li Jigang, and Boliang Zhu for their helpful feedback and support.
Declaration of Conflicting Interests
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author received no financial support for the research, authorship, and/or publication of this article.
Notes
Author Biography
Interviews
11 County P township civil affairs official 38 Cixi landless farmer 39 Cixi police officer 40 Cixi landless farmer 43 Cixi landless farmer 44 Cixi landless farmer 115 Songyuan county deputy party secretary 159 Songyuan county petitioning official 171 Ningbo district land resources official 182 Ningbo district land resources official 183 Songyuan county finance official 184 Songyuan county labor and social security official 193 Songyuan county judge 194 Songyuan county land resources official 264 County Y petitioning official
